We’re often approached by new members with questions about how to register their content. Not just about the process - for example, how to submit metadata, or construct DOI suffixes - but about which record types to use for their particular type of content.
Our metadata deposit schema supports 11 named record types. You can find an introduction to each type at the following links:
- Books, chapters, and reference works
- Conference proceedings
- Datasets
- Dissertations
- Grants
- Journals and articles
- Peer reviews
- Pending publications
- Preprints and posted content
- Reports and working papers.
- Standards
Crossref membership is open to organizations that produce professional and scholarly materials and content. The problem becomes quickly apparent: not all “professional scholarly materials and content” fits within the above named categories.
So, what do you do when you want to register content that falls outside of, or between, those record types?
As a first step: tell us. We have incrementally updated our schema over the last 25 years to add support for more record types, and more metadata elements in response to the needs of our member community. Preprints were added in 2016; peer reviews in 2017; grants in 2020. We keep track of the requests we receive for new record types, and use that in decision-making for future schema updates.
But, the truth is that schema updates are slow and painstaking work. There are a lot of considerations and planning, and changes to our system code, and metadata retrieval services that are involved in supporting a new record type. And, we want members to be able to register their content now, even if the metadata schema isn’t perfectly suited for it, rather than have to wait around until the schema catches up.
That means you’ll have to pick one of the existing record types to use, even if it doesn’t specifically name your type of content, based on whether the structure of metadata elements supported within it is close enough to what you need.
I’ll briefly touch upon some commonly asked about types of content and possible solutions.
Conference-related materials that are not strictly “proceedings”
This could be visual materials like conference posters, videos of presentations of conference papers, or abstracts of conference papers that are compiled into some parent work other than a proceedings. In general, regardless of the actual medium or content type, it’s still best to use the proceedings record type to register any materials that stem from a conference event.
When you’re submitting metadata, you can treat the posters, videos, abstracts, etc. all as if they were conference papers in a proceedings.
It’s okay if no actual parent-level object like a proceedings even exists. You don’t need to assign a DOI at the proceedings-level, and you’re only required to supply a title, publisher name, and publication year to describe the proceedings ‘container’. For example, if you were registering a collection of posters from the 2025 Annual Meeting of the Society of Metadata Idealists, you could supply the title as something like “Annual Meeting of the Society of Metadata Idealists, 2025 Poster Exhibition”; the year as “2025”; and the publisher as “Society of Metadata Idealists”.
Audio, Video, and Image content
We are often asked about registering a variety of scholarly or research-related content in audio or visual media: lectures, podcasts, maps, graphs and charts, video tutorials, etc.
If these are supplementary materials related to a journal article, book chapter, or other registered publication, you should consider registering them as Components. This automatically creates a parent/child relationship between the component and the publication it’s a component of.
If they are independent items, the best option at this point in time is to use the Posted Content record type. This record type is most commonly used to register preprints, but the schema has a ‘type’ attribute in the main <posted_content> tag, and that can be populated with any of the following:
- preprint
- working_paper
- letter
- dissertation
- report
- review
- other
Among those, the “other” type is a very useful catch-all category. When possible, “posted content (other)” (as opposed to “posted content (preprint)” ) is an excellent option for content that doesn’t fit squarely into the existing record types.
Educational materials
Sometimes universities and other institutions want to register instructional content like curricula, course materials, recorded lectures, and full courses/MOOCs. This isn’t always an easy fit with our schema. For standalone items, it’s definitely possible to use posted content (type: other) as a catch all, as explained above. But, for a full course, lecture series, or any other collection of items that you want to be reflected as part of a group, there’s another consideration. Do you want all the materials from within a course or lecture series to be grouped together under the umbrella of a larger parent/container object? If so, you may want to register them as “reports” within a report series. Or, as “datasets” within a database.
Articles that aren’t part of a journal, preprint repository, or other ‘container’
As we recently noted on our own blog, we’ve begun registering Crossref’s blog posts in the spirit of ‘eating our own dogfood’ (apologies for the unpleasant metaphor, for those who were not familiar!). There are good reasons for Crossref members to register the blog posts they produce as well.
Initially, our registration process for blog posts involved entering the metadata manually using the Web Deposit Form, and because of that they were registered using the reports record type. However, we’ve recently updated the registration process to produce and submit the xml metadata programmatically instead. So, we’ve opted to switch to the posted content (type: other) record type. That’s what we’d encourage members to do as well, not just for blog posts, but for other types of standalone articles as well.
How about “Datasets”?
The datasets record type has also been used as a kind of catch-all category, even though “datasets” themselves are quite specific. It’s not that these other types of content (images, audio, software, etc.) are similar to actual datasets in any way. Rather, the dataset record type is useful because it’s very minimalist and flexible. There is very little strictly required metadata, but the schema supports a full array of metadata optionally, including a free text <description> field, which can be particularly handy in the case of ambiguous content types.
Caveats and backup options
The posted content record type is great, but it has a couple of downsides. Crossmark isn’t supported for posted content. That’s not an issue for most members, but if you typically implement the Crossmark service and want to continue to do so for all the content you register, it’s something to consider.
You also can’t add components to posted content metadata. Again, not the biggest deal, but say you were registering a lecture, and you wanted to add DOIs for a few supplementary materials, perhaps worksheets or some slides with illustrations, that would normally be a reasonable use case for components. So, in that case, you’d have to pick a different record type for the main lecture’s DOI.
A bigger concern for both the posted content and datasets records types is that the only way to register items using those types is to construct and submit xml files directly. Our manual metadata deposit forms, and the very popular OJS plugin, don’t support them. So, what can you do if you don’t have the technical resources to create xml metadata deposits?
The new Record Registration Form only supports journal content and grants at the moment, though we have plans to build it out for more record types in the future. Journals and grants are not particularly open or flexible, so aren’t typically suitable solutions for instances where record type is unclear. The long-slated-for-deprecation Metadata Manager was only ever built to register journal content, and proved itself to be too buggy and too resource-intensive to even consider adding new record types to it.
However, the old stalwart Web Deposit Form supports several record types, and among them are conference proceedings, which can be used for any conference or event related materials, as well as reports. The reports record type was created with reports, working papers, and white papers in mind. Conceptually, there’s some overlap between a working paper and a preprint, or a report and a (non-journal) article. So, using the reports record type can be a useful way to get some of the more ambiguous or not-accounted-for-yet types of content registered without needing to create your own xml.
Relax, it doesn’t have to be perfect
In the service of wanting to support and encourage as much content registration as our members want and need, sometimes we have to think outside the self-imposed boxes of our metadata schema. The goal is to describe the content items as accurately as possible, even if that’s just ‘good enough’ and not precisely perfectly. And, as the schema continues to evolve, and new record types are added in the future, you can always update your metadata records when a record type comes around that’s a better fit.